Adding instead of subtracting the products in the loop allows the
compiler to generate more efficient multiply-accumulate instructions
when 16-bit multiply-subtract is not available. ARM has only
multiply-accumulate for 16-bit operands. In general, if only one
variant exists, it is usually accumulate rather than subtract.
In the same spirit, using the dedicated saturation function enables
use of any special optimised versions of this.
Signed-off-by: Mans Rullgard <mans@mansr.com>
"for (i = 1; i < num + 1; i++)" look like "for (i = 1; i <= num; i++)".
Programmers worldwide rejoiced and used the operand, e.g. in the thread
"[PATCH] remove ugliness in cel_filters.c", and He saw that it was good.
Originally committed as revision 20489 to svn://svn.ffmpeg.org/ffmpeg/trunk
convolution in the upcoming AMR-NB floating point decoder. The function scales
and adds a vector, that is lagged by some offset, to another vector with the
same number of elements.
Patch by Colin McQuillan ( m.niloc googlemail com )
Originally committed as revision 19634 to svn://svn.ffmpeg.org/ffmpeg/trunk
Martin Storsjö
Mans Rullgard
Mohamed Naufal Basheer
Alex Converse
Vitor Sessak
Ronald S. Bultje
Colin McQuillan
Diego Biurrun
Kenan Gillet
Reynaldo H. Verdejo Pinochet